New horizons at L’Anse aux MeadowsPaul M. Ledgera,b,1, Linus Girdland-Flinkc, and Véronique Forbesa

aDepartment of Archaeology, Queens College, Memorial University of Newfoundland, St. John’s, NL, Canada A1C 5S7; bDepartment of Geography,Memorial University of Newfoundland, St. John’s, NL, Canada A1B 3X9; and cSchool of Natural Sciences and Psychology, Liverpool John Moores University,L3 3AF Liverpool, United Kingdom

Edited by Dolores R. Piperno, Smithsonian Institution, Washington, DC, and approved June 17, 2019 (received for review May 9, 2019)

The UNESCO World Heritage site of L’Anse aux Meadows (LAM) innorthern Newfoundland is the only undisputed site of pre-1492 pres-ence of Europeans in the Americas. In August 2018, we undertookfieldwork at LAM to sample the peat bog 30m east of the Norse ruinsfor a multiproxy paleoenvironmental assessment of Norse settle-ment. Instead, we encountered a new cultural horizon. Here we re-port our fieldwork at this iconic site and a Bayesian analysis of legacyradiocarbon data, which nuance previous conclusions and suggestNorse activity at LAM may have endured for a century. In light ofthese findings, we reflect on how the cultural horizon, containingnonnative ecofacts, may relate to indigenous or Norse activities.

Norse | indigenous | Bayesian modeling | insects | pollen

The Norse colonization of the North Atlantic, a defining fea-ture of the Viking Age, reached its ultimate extent on the

shores of Newfoundland some 4 centuries before ChristopherColumbus (1). Pre-Columbian presence of Europeans in the NewWorld had often been suspected on the basis of oral historiesinscribed by mid-13th century Icelandic scholars in GrænlendingaSaga and Eiríks Saga Rauða. These sagas recounted how Norseexplorers from Greenland discovered and established a settlementin Vínland, a land west of Greenland (2). Nevertheless, it was notuntil the 1960s that definitive evidence was discovered near theNewfoundland outport of L’Anse aux Meadows (LAM) (1).Throughout the 1960s, 8 Icelandic-style turf structures were ex-

cavated at LAM (Fig. 1), identifying Norse material culture such asa bronze cloak pin, a steatite spindle whorl, and evidence of rudi-mentary iron working (2). The remainder of the European materialculture was nondiagnostic, comprising fragmentary iron, primarilynails and rivets, a bone needle fragment, and a whetstone (1).Numerous stone tools pertaining to indigenous peoples were alsorecovered (2), perhaps suggesting LAM was the theater of theearliest contact between Europeans and North Americans. In the1970s, further excavations by Parks Canada revealed iron-workedwooden objects and jasper fire starters matching rocks that outcropin western Greenland and Iceland (2). A wealth of wooden debit-age was also uncovered, which was attributed to both indigenousand Norse occupations (2). Compared with Norse sites in Green-land, the limited suite of material culture from LAM is notable forthe near absence of steatite and osseous material commonly usedfor all manner of implements in Greenland (Fig. 1). Botanicalanalyses identified wood and nuts from the White Walnut (Juglanscinerea), an exotic species in Newfoundland that suggests wider-ranging Norse voyages to the south (2).Over 100 ecofacts associated with the archaeological heritage

(wood and charcoal) were submitted for radiocarbon dating (2).Dates ranged from 5,080 ± 110 14C y B.P. (Qu-365; charcoal) to260 ± 110 14C y B.P. (S-1167; worked wood), encompassing thesuite of occupations at the site. Assays from Norse contexts (56assays) ranged from 1,630 ± 70 14C y B.P. (WAT-420; charredwood) to 865 ± 65 14C y B.P. (S-1091; wood). Icelandic sagaliterature, and references to Vínland in an ecclesiastical treatisefrom ca. AD 1075, suggested LAM should date to ca. AD 1000,something the radiocarbon data seemed to challenge (2). It wasnot until the “in-built” ages of materials such as wood andcharcoal became widely recognized that these conflicting narra-tives could be reconciled. Assays on short-lived macrofossils

(twigs) from Norse contexts dated to between 1,050 ± 65 14C y B.P.(S-1340; Abies sp. twig; Cal AD 895–1030) and 955 ± 100 14C y B.P.(S-1355; Larix sp. twig; Cal AD 995–1185), suggesting an occupa-tion centered on AD 1000 (2). A weighted mean of twig dates—notwithstanding issues associated with combination of 14C agesfrom multiple individuals—provided a result of AD 986–1022 (3).Despite being the earliest known European outpost in North

America, LAM remains enigmatic. None of the structures areidentifiable as animal shelters, nor is there faunal evidence foranimal husbandry—the foundation of Norse subsistence inGreenland and Iceland (2). LAM is therefore an outlier withinthe Norse settlements of the North Atlantic. The paucity ofmaterial culture and shallow deposits indicate a transitory placefunctioning as a base for exploration of North America (2). Al-though the Norse colonization of the North Atlantic is oftenviewed as a search for farmland, it was also an endeavor to secureluxury resources for European markets (4). From this perspective,perhaps a Norse outpost makes perfect sense. LAM is located onthe shores of a rich cod fishery—which, 6 centuries later, was hometo hundreds of seasonal French fishers (5)—and in a dense nestingregion for eiders (6). Both stockfish (dried cod) and eiderdownwere highly prized commodities in the Viking Age (4, 7).

FieldworkOur trench, measuring 0.65 × 1.50 m, was located 30 m east ofRuin D (Fig. 1). The new cultural horizon (4A800B7) was encoun-tered between 35 and 45 cm (Fig. 2) and comprised finely laminated(0.5 to 1.0 cm thick) apparently trampled surfaces containing char-coal, wood debitage, and charred plant remains. Abundant well-preserved plant (leaves and twigs) and insect remains were evident tothe naked eye. A 2-L bulk sample of the layer and a monolithspanning the cultural horizon and natural peat were collected (Fig. 2).

Environmental Archaeological AnalysesThe peat monolith was cut into contiguous 1 cm3 subsamples,and the pollen content was analyzed (e.g., ref. 8). A 450-mLsubsample of the 2-L bulk was disaggregated in <2% NaOH,then washed through a series of sieves. Residues were hand-sorted under a binocular microscope to isolate charcoal, seeds, in-sects, and wood debitage. Short-lived plant macrofossils selected from3 levels (as illustrated in Fig. 2) were submitted for C14 dating at theAndré E. Lalonde Accelerator Mass Spectrometry Laboratory.The pollen content is notable for tree, shrub, and heath percent-

ages (in particular,Myrica), which are high in respect to the aspect ofthe site and previous studies at LAM (1, 9). Apophytes (e.g., Rumexsp. Achillea millefolium type) are elevated at ca. 10% total landpollen (TLP), while exotics such as Juglans (Walnut) and Humulustype (hops or cannabis) are also present. Sporormiella-type fungi,exclusively associated with the dung of grazing herbivores (caribou inthe case of Newfoundland), were also found in 7 of 12 samples.

Author contributions: P.M.L. and V.F. designed research; P.M.L., L.G.-F., and V.F. per-formed research; P.M.L. and V.F. analyzed data; and P.M.L. and V.F. wrote the paper.

The authors declare no conflict of interest.

This open access article is distributed under Creative Commons Attribution License 4.0 (CC BY).1To whom correspondence may be addressed. Email: pledger@mun.ca.

www.pnas.org/cgi/doi/10.1073/pnas.1907986116 PNAS Latest Articles | 1 of 3

ANTH

ROPO

LOGY

BRIEFRE

PORT

Dow

nloa

ded

by g

uest

on

May

18,

202

0

The concentration of insect remains in 4A800B7 (0.24 beetlesper mL−3) is very high compared with natural peatlands. Mostabundant are rove beetles such as Pycnoglypta sp., typically domi-nant on archaeological sites. Other examples include Acidotaquadrata (Zetterstedt), a Holarctic species previously unrecorded inNewfoundland, and Simplocaria metallica (Sturm), a pill beetleconsidered adventive (nonnative) in Canada (10). Of further note isa dock/sorrel seed that compares favorably to Rumex aquaticus L., aspecies of uncertain status in Newfoundland (11).

Bayesian Modeling of Radiocarbon DataLegacy 14C data pertaining to the Norse and indigenous (often re-ferred to as Recent Indian, a cultural group ancestral to the Beothuk)occupations at LAM were collated from the Canadian Archaeologi-cal Radiocarbon Database and placed within a relative chronologicalframework (Fig. 2) using published data (1, 2). A series of Bayesianmodels were constructed in OxCal (Fig. 2F) to estimate start and enddates for the Norse and indigenous occupations, individual structures,and 4A800B7 under varying prior information. Modeling results wereremarkably consistent (Fig. 2F), and model A suggests Norse occu-pation began Cal AD 910–1030 (2σ; boundary: Start Norse), endedCal AD 1030–1145 (2σ; boundary: End Norse), and endured for 0 to195 y (2σ; span: Sequence Norse). Greater uncertainty surroundsindigenous occupations, where a start of Cal AD 710–1130 (2σ;boundary: Start pre-Beothuk) and end of Cal AD 1540–1815 (2σ;boundary: End pre-Beothuk) is indicated. Assays from 4A800B7 dateit to the mid-12th to the late 13th centuries (Fig. 2).

DiscussionBayesian modeling has confirmed previous conclusions (2, 3)regarding Norse arrival, but suggests a potentially longer than as-sumed period of use (up to 195 y). This does not imply a continuousoccupation, which, given the shallow cultural deposits, seems un-likely. Rather, it indicates the possibility of sporadic Norse activitybeyond the early 11th century. Data from indigenous contexts is lessprecise, and activity is modeled to have begun between the 8th and

12th centuries. LAM therefore could have been a shared zone ofinteraction. Radiocarbon data place 4A800B7 in the 12th and 13thcenturies, implying an association with indigenous activities.Assuming an indigenous association poses the question: What

process formed 4A800B7? The deposit is thick (≥10 cm), withexcellent preservation and fine laminations, which indicate rapidaccumulation and burial of trampled material. The concentrationsof insects, wood debitage, charcoal, and weed pollen are suggestiveof abundant organic waste. Sporormiella-type fungi also point to theincorporation of dung into the matrix of the deposit, or the nearbypresence of grazing animals. Similar deposits are well documentedfrom Greenlandic Norse sites (8), but contrast with shallow culturallayers from contemporary indigenous settings in Newfoundland.The ecofactual content of 4A800B7 poses interesting bio-

geographical questions, but offers few clues to its cultural affinity. Thedock seed (cf. Rumex aquaticus) may derive from an introduced plantnative to Eurasia, but the tendency of Rumex species to hybridizecomplicates secure identification. Cereal-type pollen could derive fromwild grasses, whileHumulus lupulus-type and Juglansmay have arrivedthrough wind dispersal. However, their sole presence in the culturalhorizon and the past identification of Juglans macrofossils at LAM isintriguing. Of the beetles, S. metallica is considered native toGreenland, where it has been found in Norse and Pre-Inuit contexts,while A. quadrata, never previously identified from Newfoundland, iscommon in the circumpolar north. If any of these species truly areintroductions to Newfoundland, their arrival by the 13th century mayhave been via either Norse or indigenous trade or migration routes.

ConclusionsThe results presented here pose more questions than answers. Thecultural deposit that we have discovered at LAM exhibits affinities withcultural layers from across the Norse North Atlantic, but our analysesdemonstrate it postdates even the most generous estimates for Norseactivity in Newfoundland. Ecofacts from cultural waste sandwichedbetween layers of peat may not be as evocative as artifacts such as aringed bronze pin, or a finely crafted lithic projectile point. Yet, they

Eastern Settlement

ICELAND

NEWFOUNDLAND

WesternSettlement

L’Anse aux Meadows

GREENLAND

LABRADOR

1960s197319741975

19762018 trench location

Excavations

Bone (undiff)

Wood

WoolLeather

Iron

Stone

Whale bone

Walrus ivory

Copperalloy

Caribouantler

Steatite

Ironslag

0

10

1

100

0.1Fig. 1. The North Atlantic region. (Left Inset)Summary of material culture from church farms(blue), farms (red), and shielings (green) in Green-land in terms of raw materials. The material cultureassemblage from LAM is represented in yellow(Right Inset). The location of Norse ruins and exca-vations at LAM.

2 of 3 | www.pnas.org/cgi/doi/10.1073/pnas.1907986116 Ledger et al.

Dow

nloa

ded

by g

uest

on

May

18,

202

0

present new horizons for examining the environmental legacies ofinter- and intracontinental movement of people within North Americaprior to 1492. Whatever their affinity, the cultural deposits withinthe LAM peatlands preserve unique archaeological and bio-geographical stories waiting to be told.

ACKNOWLEDGMENTS.We thank Birgitta Wallace for stimulating discussionsand support and the Parks Canada staff at LAM and Dartmouth. We alsowish to thank Meghan Burchell and Bryn Tapper for help with imageryand Dawn Mooney and Yannick Miras for advice on botanical IDs.We acknowledge support from Memorial University Faculty Start-UpFunds.

1. A. S. Ingstad, Excavations of a Norse Settlement at L’Anse Aux Meadows, New-foundland 1961-1968 (The Norse Discovery of North America, Norwegian UniversityPress, Oslo, Norway, 1985), Vol. 1.

2. B. L. Wallace, The Norse in Newfoundland: L’Anse aux Meadows and Vinland. Nfld.Stud. 19, 5–43 (2003).

3. R. Nydal, A critical review of radiocarbon dating of a Norse settlement at L’Anseaux Meadows, Newfoundland, Canada. Radiocarbon 31, 976–985 (1989).

4. C. K. Keller, Furs, fish and ivory: Medieval Norseman at the Arctic fringe. J. NorthAtlantic 3, 1–23 (2010).

5. B. Tapper, “An archaeological analysis of the distribution of French fishing rooms onthe Petit Nord, Newfoundland,” Masters thesis, Memorial University of Newfound-land, St. John’s, NL, Canada (2014).

6. T. J. Kristensen, J. E. Curtis, Late Holocene hunter-gatherers at L’Anse aux Meadows and thedynamics of bird andmammal hunting in Newfoundland.Arctic Anthropol. 49, 68–67 (2012).

7. V. Forbes, Duck fleas as evidence for eiderdown production on archaeological sites.

J. Archaeol. Sci. 61, 105–111 (2015).8. P.M. Ledger, K. J. Edwards, J. E. Schofield, Taphonomy or signal sensitivity in palaeoecological

investigations of Norse landnám in Vatnahverfi, southern Greenland? Boreas 44, 197–215

(2015).9. A. M. Davis, J. H. McAndrews, B. L. Wallace, Paleoenvironment and the archaeo-

logical record at the L’Anse aux Meadows site, Newfoundland. Geoarch 3, 53–64

(1988).10. Y. Bousquet, P. Bouchard, A. E. Davies, D. Sykes, Checklist of Beetles (Coleoptera) of

Canada and Alaska (Pensoft, Sofia, Bulgaria, 2013).11. S. J. Meades, L. Brouillet, “Annotated checklist of the vascular plants of Newfoundland

and Labrador.” Flora of Newfoundland and Labrador (2019). https://newfoundland-

labradorflora.ca/index.cfm. Accessed 9 May 2019.

25 μm

10 mm

South West North East

4A800B7-1

Smithy ActivityT-393; S-1101; T-366; S-1127

House Aoccupation

T-310

FirepitS-1124 Start of Norse occupation

House D occupation

WAT-420; T-364

Construction

House F occupationT-325; T-326; S-1351;

S-1108; S-1350; St-2665?

Start House D

End House D

T-817

End House AEnd Smithy

ConstructionT-531; S-1104; S-1102

End House F

Other Norse activity

T-327; T-324; T-306; T-309;

T-367

End of Norse occupation

Bog/MiddensTO-116; S1092; S-1119; S-1358;S-1339; S-1346;S-1109; S-1343; S-1348; S-1090; S-1118; S-1115; S-1342; S-1347; S-1106; S-1352; S-1357; S-1114; S-1112; S-1120; S-1110; S-1093; ; ; ; S-1113 TO-119 S-1111

S-1091; ; ; ; ; S-1340 TO-117 S-1355TO-118Qu-350; WAT-436

800B Activity

UOC-7390UOC-7389

End of 800B

Start of 800B

UOC-7388

Start House FStart House AStart Smithy

ConstructionT-530; T-818

Dark brown to black poorly humified ericaceous PEATwith patches of Polytrichum

Surface vegetation and root mat.

Dark brown to black poorly humified PEAT rich in Polytrichum and monocot remains

Orange brown to golden brown poorly humified Sphagnum PEAT.

Red brown (becoming black brown) laminated PEAT rich in poorly humified woody remains, monocots, Sphagnum, charcoal and insects.

Orange brown to golden brown poorly to moderately humified SphagnumPEAT.

33Dep

th

20

Betula

90

Myrica

gale

50

Picea

Pinus

50

Ericac

eae

Alnus

20

Rubus

cham

aemoru

s

20

Poace

ae

Cypera

ceae

Sangu

isorba

Apiace

ae

20

Cheno

podia

ceae

Rumex

aceto

sa-ty

pe

Rumex

aceto

sella

Ranun

culus

acris

-type

Lactu

ceae

Achille

a-typ

e

Brassic

acea

e

Montia

fonta

na

Asterac

eae

Sordari

a-typ

e (55

A)

Sporor

miella-

type (

T-113

)

Podos

pora-

type (

T-368

)

Gelanis

pora

(T-1a

)

Humulu

s lup

ulus-t

ype

Cereal-

type

Jugla

ns

0 100

Trees

, Shru

bs an

d Hea

ths

HerbsApo

phyte

s

Exotic

s

Trees, shrubs and heaths Herbs Apophytes Fungi Exotics

45

S-1167; Qu-348; T-410;S-1361; Qu-352; S-1166;S-1354; S-1129; Qu-349;

Beta-252633; Beta-252632; T-411; S-1123; S-1098;S-1099; T-365; S-1095;T-368; S-1121; S-1126

Pre-Beothuk activity

Start pre-Beothuk

End Pre-Beothuk

1 mm

1 mm

500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

UOC-7388 (633 ± 23; Sphagnum moss; 36-35 cm)

UOC-7389 (733 ± 22; Larix sp. twig; 38-37 cm)

UOC-7390 (847 ± 22; Indet. twig; 45-44 cm)

Boundary: End Norse

Boundary: Start Norse

Boundary: End pre-Beothuk

Boundary: Start pre-Beothuk

Modelled date (cal. AD)

0 100 200 300 400 500 600

Model AModel BModel CModel DModel EModel F

A

B

C D

E

HDuration (years)

F

G Span Duration of pre-Beothuk 630-895 [1σ]; 475-1045 [2σ]

700 800 900 1000 1100 1200

Span Duration of Norse 0-100 [1σ]; 0-195 [2σ]

Fig. 2. Results of analyses. (A) Trench stratigraphy and monolith sampling location. (B) Percentage pollen diagram displaying selected taxa (sum = ≥500 TLP),“+” indicates <1% TLP. (C) Insects and seeds from 4A800B7 (left to right): Eanus macullipennis, S. metallica, A. quadrata, Pycnoglypta sp., and dock seed (cf. R.aquaticus). (D) Pollen (left to right): H. lupulus-type, Juglans, and cereal-type. (E) Wood debitage from 4A800B7. (F) Probability distributions for start and endof Norse and indigenous occupations and calibrated C14 assays from 4A800B7. (H) Modeled durations for Norse (red) and indigenous (blue) activity. (G) Priorinformation incorporated into all Bayesian models (A–F). Dates in italics were treated as outliers using the OxCal charcoal outlier model. Assays in bold red areshort-lived elements also treated as outliers in models B, D, and F. Terminus post quem of 890 ± 60 in models C and D and 985 ± 1 in models E and F wasincluded to indicate that Norse settlement at LAM occurred after the settlement of Iceland and Greenland, respectively.

Ledger et al. PNAS Latest Articles | 3 of 3

ANTH

ROPO

LOGY

BRIEFRE

PORT

Dow

nloa

ded

by g

uest

on

May

18,

202

0